The goal of our project is to conduct a population-level investigation of social contexts, stress, and health. We integrate multiple biomarkers of stress with current theory and measurement tools from the social and behavioral sciences to gain a better understanding of how social contexts "get under the skin" to affect stress and health in young adults. We use data from four waves of the National Longitudinal Study of Adolescent Health (Add Health) in pursuit of three objectives: 1) Examine the interrelations among multiple subjective and physiological measures of stress; 2) Investigate how measures of socio-economic status, neighborhood quality, and interpersonal relationships in childhood/adolescence and over the transition to adulthood influence stress in early adulthood; and 3) Test competing theoretical models of the joint influence of socio- economic status, neighborhood quality, and interpersonal relationships on stress. We examine subjective measures of stress, including perceived stress, symptoms of depression, and symptoms of anxiety, as well as physiological measures of endocrine, cardiovascular, metabolic, immune, and inflammatory activity. We take advantage of the breadth and depth of the Add Health dataset to conduct the most comprehensive investigation to date of how social stress affects physical and mental health. Additionally, the large sample size and racial, ethnic, and economic diversity of Add Health allow us to investigate the extent to which stress leads to health disparities. Stress is a major contributor to morbidity and mortality, and investigating the physiological mechanisms through which social contexts affect human health can help investigators, clinicians, and policy makers to: a) identify adverse psychosocial environments; b) better understand predisease and disease pathways; and c) inform interventions to reduce exposure to stressors, reduce the adverse impact of stressors on physiology, and prevent the emergence of stress-related disorders. [unreadable] [unreadable] [unreadable] [unreadable]